Replication packages for each accepted manuscript are stored in a ZIP file on Dropbox, with mostly consistent naming conventions. These ZIP files are created manually and uploaded to the publisher's website.
All packages stored on https://zenodo.org/ with metadata and DOI so that they are citable. Those packages that are small enough should also be on https://github.com/restud-replication-packages
- As a reader, I want to be able to browse replication packages easily, so that I get inspired for doing similar work.
- As a data editor, I want to accept GitHub repos as submissions so that I can encourage tech savvy authors to submit.
- As a production manager, I want easy upload to our repos so that I don't make any mistakes during manual work with ZIP files.
- As an author, I want the submission process to be smooth so that I can focus on my research.
- As a reader, I want to have precise citation (inlcuding DOI) easily availabe so that I can give credit to researchers.
- Explore Github and Zenodo API.
- Explore folder and file structure in Restud backlog.
- Build a datastore of metadata of replication packages. (zenodo UI? submisssion form in .docx format? new, structured submission form?)
- Develop/test tool to upload ZIP files to GitHub as new repos.
- Develop/test tool to pull these repos over to Zenodo and publish them with a DOI.
- Scipt prepares packages and metadata and stores them in a staging area
- User manually checks and edits all metadata and approves packages. UI: .csv file or Google Spreadsheet
- request edits from authors?
- Script requests DOI for each package and saves them in metadata in staging area
- Script uploads small packages to github (creating a release) and zenodo
- Script uploads large packages to zenodo, entering a simple README.md with metadata, doi and link to zenodo on github
- Author filles in sctructured submission form. UI: jotform or similar
- Metadata and package enter staging area (what backend to use?)
- Data Editor reviews package
- Script uploads to github and zenodo